The management of patients with colorectal liver metastases is still controversial. Recent evidence suggests benefit for resection of localized metastases within a single lobe of the liver. A series of 49 patients undergoing liver resection for localized liver metastases is presented. Resection involved right hepatectomy (23), left hepatectomy (8) and segmental excision (18). The median age was 62 (range 40-77). Patients have been followed for a period of up to seven years. The overall median survival was 24 months. The actuarial 3-year survival was 57.9%. There was no significant correlation of survival with the degree of differentiation or Dukes staging of the primary tumour. This series adds further weight to the evidence supporting resection in selected patients with colorectal liver metastases.

Peri-tumour fibrin is a consistent feature of tumour stroma and is deposited shortly after tumour cell inoculation. Since there are several ways in which fibrin may be beneficial to tumour growth, it is possible that the ability of normal or malignant tissue to generate fibrin may influence metastasis. Many normal tissues and tumour cells possess a procoagulant activity that is due to a complex of tissue factor and factor VII. We have measured this tissue procoagulant activity in normal rats, rats stabilised on Warfarin and similarly anticoagulated animals injected with factor VII. The effect of Warfarin and factor VII administration on pulmonary seeding following injection of MC28 fibrosarcoma cells was also assessed. Procoagulant activity in adrenal, lung and colon was significantly reduced by Warfarin (P less than 0.001). Administration of factor VII significantly increased lung and adrenal tissue procoagulant activity in anticoagulated rats (P less than 0.02). Warfarinised rats had significantly slower primary tumour growth (P less than 0.001) and fewer lung deposits than control animals (P less than 0.001). Injection of factor VII restored pulmonary seeding to control levels (P less than 0.001). Warfarin did not affect the ability of the cells to adhere in vitro and did not reduce the number of tumour cells physically trapped in the lungs after intravenous injection. It is concluded that the procoagulant activity of normal tissues may influence their ability to support tumour growth and that the antimetastatic effect of Warfarin may be at least partly due to a reduction in the availability of the factor VII required for this activity.

Procoagulant activity (PCA) in normal urine has been recognized for over 50 years. Although tissue factor (TF) is produced by certain tumours, and is increased in both tumour-associated macrophages and blood monocytes, the possibility that it might also be increased in urine has not been studied in patients with cancer. We have measured urinary PCA in hospital controls without inflammatory or neoplastic disease (n = 79), in patients with rheumatoid arthritis (n = 8), inflammatory bowel disease (n = 19), colorectal cancer (n = 70) and in patients undergoing colonoscopy (n = 50). Urinary PCA was higher (P less than 0.001) in patients with colorectal cancer and inflammatory bowel disease than controls or patients with rheumatoid arthritis. Fourteen (88 per cent) out of 16 colonoscopy patients subsequently found to have carcinoma or inflammatory bowel disease had levels above the control upper quartile, compared with 8 (24 per cent) out of 34 with normal colonoscopy (P less than 0.001). TF inhibitors confirmed the nature of the PCA and Western blotting studies indicated a urinary TF molecular weight of approximately 38,000. These studies provide further evidence of abnormal haemostasis in malignancy and suggest that determination of urinary TF may provide a useful screening test in patients undergoing colonoscopy.